Confused about the number of houses Ireland needs to build each year? You aren’t the only one. Estimates range from as low as 33,000 to as high as 80,000-plus. How on earth do we make sense of this? And what does it mean for policy in this vital area? There are two specific reasons why it is so difficult to calculate housing demand and why it is vital to understand what different estimates refer to.
The background
It has been clear for some time that the targets in the Government’s Housing for All plan, averaging an extra 33,000 additional homes each year, were out of date. Since the plan was published the estimates for the recent growth rate of the population has increased – thus increasing the housing need. But how much higher should the figure be?
The Government looked for input on this issue from the Economic and Social Research Institute (ESRI), as part of work on a revision of a wider National Planning Framework for the State. It also established the Housing Commission, which considered the building needs in the years ahead as part of its work. And other forecasters have also been putting their oars in.
Both the ESRI and the commission published forecasts in ranges, rather than precise figures. The ESRI said that based largely on different assumptions on inward migration to the State, housing needs could be as low as 35,000 per annum or as high as 53,000. Its paper pointed out that an average figure from its work would be about 44,000. The commission had an even wider range, from 33,400 a year to 81,400, though its estimates were drawn up on a slight different basis, which we will look at below. Meanwhile, Davy stockbrokers said in a recent report that as many as 85,000 new homes might be needed each year.
If our finances go flat, how will Ireland pay its bills?
The key decisions now facing Donald Trump which will have a big impact on the Irish economy
Election campaign got off to spluttering start on economic and budget issues. Here is what it all means
Donald Trump’s economic plan: the key areas for Ireland
Let’s look at the two big reasons why researchers are coming up with such a variety of numbers.
Uncertainty over immigration
Housing needs depend on the population size. Here, forecasting focuses in large part on trends in births and deaths that change slowly over time and are relatively predictable, within a certain margin of error. However, there is another, much-less predictable factor – net migration. This refers to the balance between emigration and immigration, which has had a vital impact on the Irish population over the years. For example in the early 2000s, the Celtic Tiger era, the Irish population surged – particularly immigration from new eastern European EU members such as Poland and the Baltic States after 2004. By 2011, one in eight people living in Ireland were not born in the country, more than double the figure in 2002. The financial crash led to this trend being reversed, but net inward migration has resumed in recent years, accelerated over the past two years by refugees from Ukraine.
Immigration is difficult to predict – and not just because of factors such as the Ukraine war. Different estimates are a key reason for the range of new home needs put forward by the ESRI, for example. Its researchers point out that by 2040 there will be a 400,000 difference in the size of the Irish population between its highest inward migration scenario and its lowest. The Davy research, meanwhile, is based on a rapid economic expansion taking the population size from about 5.3 million now to 5.9 million by the end of this decade, well ahead of other estimates, driven by strong inward migration to fill available jobs. A key question is the extent to which the housing shortage might constrain such an inflow.
These different migration assumptions are key reasons for the different estimates of housing needs – and why researchers typically come out with a range of forecasts. There are also other factors in the mix, in looking at what is called structural housing demand, in other words the additional demand for housing related to future population growth.
One is the average size of households. At about 2.74 in 2022, average household size in Ireland is high by international standards and can be expected to fall. How quickly is the question.
Another important factor is obsolescence rates – each year a certain amount of the housing stock in effect wears out and needs to be replaced. Estimating the rate at which this happens – and the rate at which currently vacant property might be brought on line – is also important.
The housing backlog
There is another key factor in the mix. Because of the low level of building after the financial crash, a big lack of supply in the market has left many young people stuck living at home. The ESRI figures do not attempt to take account of this backlog factor, saying this issue was outside the scope of its latest study, which was set up to look at the impact of future population growth and demand. However, to the extent that the report assumes that average household sizes will fall, it does take some limited account of this backlog.
The commission did try to include the backlog in its estimates and said it would lead to a very significant increase in annual housing targets and the requirement to “take exceptional and radical measures to deliver a substantial amount of housing in the shortest time practicable”. It estimated the backlog figures at about 235,000 houses on the basis of its view that housing size here, if there was an adequate supply, would average about 2.4 to 2.45 and would fall sharply in the years ahead.
There have been a range of estimates of this backlog – or deficit, as the commission calls it – and by its nature there is no way of arriving at an exact figure. Generally researchers look at international trends in house size and at the slowing in reduction in this figure since 2011, the year when house building fell off a cliff.
The debate
The ESRI forecasts have set off something of a debate, which underlines the key issues. Two former members of the commission – Dermot O’Leary of Goodbody and Dr Ronan Lyons of TCD – have said that in their view the ESRI’s predictions – centring on an annual housing need of about 44,000 – are on the low side. This results in part from the ESRI not fully accounting for the historical housing deficit and its related assumption on household size.
Lyons said that updating the commission research to today would likely increase the housing deficit or backlog figure from the 235,000 estimate in the report to about 300,000 today. On X, he said that “by setting the housing deficit to one side, the ESRI have misunderstood the single most important parameter in estimating housing requirements to 2040”.
O’Leary‚ in a note to clients, also highlighted this point, while, along with Lyons, he felt that the ESRI’s migration estimates are “well below recent trends”.
ESRI researcher Dr Adele Bergin, one of the authors of the report, said the goal of the report was to look at what is called “structural demand” – which is future demand based on population growth and other trends and that this was clearly spelt out in the document. The deficit issue would also need to be considered in setting national targets. She said the report’s analysis on the controversial issue of household size was based on previous ESRI work which found that as well as the housing shortage, the current relatively large size of Irish households also reflected the relatively young Irish population by European standards and the relatively high fertility rates here. In other words, estimating the underlying trend here if there was ample supply is not straightforward.
The fallout
The ESRI research is due to feed into the new National Planning Framework which will look at where houses and associated infrastructure need to be built. The Climate Advisory Council recently underlined the urgency of completing this and the need to accelerate building where there are good public transport links, shops, schools and so on to cut car journeys. The Government needs to absorb all this and set its new housing targets, with both the ESRI and the commission underlining that whatever figure or range is set will need to be adjusted over time in line with trends.
But where to start? A strange aspect of the commission report was that its range of annual housing requirements was very wide – from 33,400 to 81,400. It was odd it did not narrow this down to give a better guide to policymakers. A straight midpoint is about 50,000 but the tone of the commission report was that it felt the needs were more likely to be on the higher side. Notably, 50,000 per annum was the figure mentioned by Minister for Housing Darragh O’Brien as a likely target in the wake of the ESRI findings.
Whatever is chosen will be highly political and feed into the next general election campaign. Two points are key. One is that as well as the overall target, it is vital that the right houses are built in the right places – for example, taking into account the need to cut emissions and deal with many more smaller households. The second is that it is all about delivery. In the campaign to come the electorate would be well advised to make their judgment on who has a coherent plan and can actually implement it, rather than who promises to “deliver” the highest number. In the policy areas, the Housing Commission is a good place to start.
